SCOS 02/2 Scientific Advice on Matters Related to the Management of Seal Populations: 2002

Contents

SCOS 02/2 Scientific Advice

SCOS 02/2 ANNEX I The Status of Grey and Common Seals in the North Atlantic Region

SCOS 02/2 ANNEX II The Status of British Grey Seal Populations Appendix 1 Estimates of pup production

SCOS 02/2 ANNEX III The Status of British Common Seal Populations

SCOS 02/2 ANNEX IV Population dynamics of grey seals

SCOS 02/2 ANNEX V Information papers provided for SCOS

1 SCOS 02/2 Scientific advice

Background Under the Conservation of Seals Act 1970, the Natural Environment Research Council (NERC) has a duty to provide scientific advice to government on matters related to the management of seal populations. NERC has appointed a Special Committee on Seals (SCOS) to formulate this advice so that it may discharge this statutory duty. Terms of Reference for SCOS and its current membership are given at the end of this document. Formal advice is given annually based on the latest scientific information provided to SCOS by the Sea Mammal Research Unit (SMRU – a NERC Collaborative Centre at the University of St Andrews). SMRU also provides to government scientific review of applications for licences to shoot seals, and information and advice in response to parliamentary questions and correspondence. This report provides scientific advice on matters related to the management of seal populations for the year 2002. It begins with some general information on British seals, gives information on their current status, and addresses specific questions raised by the Scottish Executive Environment Rural Affairs Department (SEERAD). Appended to the main report are two annexes giving more detail about the status of the two species of seal around Britain: grey and common (harbour) seals.

General information on British seals Grey seals The grey seal is the larger of the two species of seal that breed around the coast of the British Isles. It is found across the North Atlantic Ocean and in the Baltic Sea (Annex I, Table 1). There are two centres of population in the North Atlantic; one in the region of Nova Scotia and the Gulf of St Lawrence and the other around the coast of the UK and especially in Scottish coastal waters. The largest population is in Canada. Populations in all three centres are increasing, although numbers are still relatively low in the Baltic where the population went through a long-term decline possibly caused by reproductive failure due to pollution. Grey seals come ashore on remote islands and coastlines to give birth to their pups in the autumn, to moult in spring, and at other times of the year to haul out between trips to forage for food at sea. Female grey seals give birth to a single white-coated pup, which moults and is abandoned by its mother about 3 weeks later. About 38% of the world population of grey seals is found in Britain and over 90% of British grey seals breed in Scotland, the majority in the Hebrides and in Orkney (Annex I, Table 1). There are also breeding colonies in Shetland, on the north and east coasts of mainland Britain and in southwest Britain. Although the number of pups born at colonies in the Hebrides has remained approximately constant since 1992, the total number of pups born throughout Britain has grown steadily since the 1960s when records began. In 2001, there were an estimated 42,000 grey seal pups born in Britain. This equates to a total population of about 130,000 grey seals. Adult male grey seals may weigh up to 350 kg and grow to over 2.3 m in length. Females are smaller at a maximum of 250 kg in weight and 2 m in length. Grey seals are long-lived animals. Males will lives for over 20 years and begin to breed from about age 10. Females often live for over 30 years and begin to breed at about age 5. Grey seals feed mostly on fish that live on or close to the seabed. The diet is composed particularly of sandeels, whitefish (cod, haddock, whiting, ling), and flatfish (plaice, sole, flounder, dab) but varies seasonally and from region to region. Food requirements depend on the size of the seal and oiliness of the prey but an average figure is 7 kg of cod or 4 kg of sandeels per day.

2 SCOS 02/2

Common seals Common seals are found around the coasts of the North Atlantic and North Pacific from the subtropics to the Arctic. Common seals in Europe belong to a distinct sub-species which, in addition to the UK, is found mainly in Icelandic, Norwegian, Danish, German and Dutch waters. Britain holds approximately 40% of the world population of the European sub-species (Annex I, Table 2). Common seals are widespread around the west coast of Scotland and throughout the Hebrides and Northern Isles. On the east coast, their distribution is more restricted with particular concentrations in the Wash, Firth of Tay and the Moray Firth Between 1996 and 2001, 34,625 common seals were counted in the whole of Britain, of which 30,196 (87%) were in Scotland and 4,245 (13%) were in England. The total British population cannot be estimated accurately because it is not possible to count all individuals in the population. Accounting for those animals that are not seen during surveys using a conversion factor leads to an estimate for the total British population of approximately 50-60 thousand animals. The population along the east coast of England (mainly in The Wash) was severely affected by the phocine distemper virus (PDV) epidemic in 1988. Numbers in England have increased since then, but have only recently recovered to the pre-epidemic level. The new epidemic in 2002 is likely to have a similar impact to the one in 1988. Common seals come ashore in sheltered waters typically on sandbanks and in estuaries but also in rocky areas. They give birth to their pups in June and July and moult in August. At other times of the year, common seals haul out on land regularly in a pattern that is often related to the tidal cycle. Common seal pups are born without a white coat and can swim almost immediately. Adult common seals typically weigh about 80-100 kg. Males are slightly bigger than females. Like grey seals, common seals are long-lived with individuals living up to 20-30 years. Common seals normally feed within 40-50 km around haul out sites. They take a wide variety of prey including sandeels, whitefish, herring and sprat, flatfish, octopus and squid. Diet varies seasonally and from region to region. Because of their smaller size, common seals eat less food than grey seals, perhaps 3-5 kg per day depending on the prey species.

3 SCOS 02/2 Current status of British grey seal populations The number of pups born in a seal population can be used as an indicator of the size of the population. Each year, SMRU conducts aerial surveys of the major grey seal breeding colonies in Britain to determine the number of pups born (pup production). These sites include about 85% of the number of pups born throughout Britain. The total number of seals associated with these regularly surveyed sites is estimated by applying a population model to the estimates of pup production. Estimates of the total number of seals at other breeding colonies that are surveyed less frequently are then added in to give an estimate of the total British grey seal population. Further details are given in Annex II and Annex IV. Pup production The total number of pups born in 2001 at all annually surveyed colonies was estimated to be 36,920. A further 5,000 pups were likely to have been born at other scattered sites. Regional estimates were 2,938 in the Inner Hebrides, 12,325 in the Outer Hebrides, 17,523 in Orkney, and 4,134 at North Sea sites. Pup production and total population size estimates for the main colonies surveyed in 2001 Location 2001 pup Change in pup Change in pup Total 2001 production production production population (to from 2000- from 1997- nearest 1000) 2001 2001 Inner Hebrides 2,938 -8.8% -0.5% 9,000 Outer Hebrides 12,325 -8.0% +1.4% 38,000 Orkney 17,523 +9.6% +4.4% 54,000 Isle of May + Fast 2,253 -10.4% +3.3% 7,000 Castle All other colonies 3,500 11,000 Total (Scotland) 38,539 119,000

Donna Nook 634 +2.6% +14.5% 2,000 Farne Islands 1,247 +6.5% -1.7% 4,000 SW England & 1,500 5,000 Wales (last surveyed 1994) Total (England) 3,381 11,000 Total (UK) 41,920 0.0%* +2.8% 130,000

*Annual change in pup production calculated from annually monitored sites only

Trends in pup production Between 1984 and 1996, estimates of the total number of pups born at regularly surveyed colonies increased year on year (Annex II, Figures 2 & 3). A small decline was observed in 1988 that was probably related to the effect of a phocine distemper epidemic that mainly affected common seals at that time. Following this event, estimated total pup production did not show any decline until 1997. It recovered again in 1998 then declined at all major breeding colonies in 1999. The greatest decline was at the Farne Islands, where pup counts are made by National Trust staff on the ground. That declines occurred at the Farne Islands and at other sites where pup production is monitored by aerial survey suggested that this was a general phenomenon and not related to differences in methods or

4 SCOS 02/2 survey conditions from previous years. Estimated pup production then increased in 2000 at all regularly surveyed breeding sites. In 2001 the pup production at regularly surveyed sites was similar to that observed in 2000. This reflects a situation in which increases in some regions, such as Orkney were balanced by declines in others, such as the Hebrides. Calculating rates of change over longer periods is more indicative of recent trends at major breeding sites. It shows, for example, that pup production in the Hebrides has not changed significantly over the past 5 years and that the average annual increase in pup production at the main breeding sites between 1997 and 2001 was 2.8%. This compares with 5.2% between 1992 and 1996 and 6.2% between 1987 and 1991(see Annex II). There is, therefore, some evidence for a slowing in pup production in recent years. The total number of pups born is the sum of pup production at many individual colonies, and because this varies from year to year, fluctuations in total pup production should also be expected. However, declines in pup production in 1999 appeared to be too great and too widespread to be explained solely by changes in survival and pregnancy rates related to shortage of space at breeding colonies.

Population size The estimated size of the UK grey seal population at the start of the 2001 pupping season was about 130,000 individuals. Taking account of uncertainty, this could range from less than 112,000 to as many 147,000 seals. The majority of these, approximately 91.5% (119,000), are associated with breeding colonies in Scotland and the remainder, 8.5% (11,000), with colonies in England and Wales.

Uncertainty in estimates There is considerable uncertainty associated with the total population estimates provided in the table above. Each estimate of pup production could lie within a range of –14% to +13% of the values provided and there are similar levels of uncertainty associated with other factors used to calculate total population size. In addition to this, the estimates of total population size assume that males have similar demography to females. Male mammals normally die younger than females. Therefore, depending upon the degree to which this is true in grey seals, the estimates of total population size in the table may be greater than the actual population size.

Trends in population size The average rate of increase in the grey seal population associated with annually monitored sites over the past 5 years (1997-2001), is +5.6%. Overall, this does not represent a significant change in the rate of increase compared to the previous 5 years (1991-1996). Even if pup production is stabilising in some regions and if these recent changes in pup production continue, the grey seal population as a whole is likely to continue increasing. This is especially the case if the observed changes in pup production are caused by reduced pregnancy rate in the seals rather than by an increase in the death rate of adults.

Current status of British common seal populations Each year SMRU carries out surveys of common seals during the moult in August. It is impractical to survey the whole of the coastline every year but current plans by the SMRU are to survey the whole coastline across 5 consecutive years. Seals spend the largest proportion of their time on land during moult and they are therefore visible to be counted in the surveys. Most regions are surveyed

5 SCOS 02/2 by a method using thermographic, aerial photography to identify seals along the coastline. Conventional photography is used in the Wash. Additional surveys using visual counts are conducted annually in the Inner Moray Firth by the University of Aberdeen.

The estimated number of seals in a population based on most of these methods contains considerable uncertainties. The largest contribution to uncertainty is the proportion of the seals not counted during the survey because they are in the water. We cannot be certain what this proportion is and it is likely to vary from region to region and in relation to factors such as state of the tide and weather. Efforts are made to reduce the effect of these factors by standardising the weather conditions and always conducting surveys within 2 hours of low tide. About 40% of common seals are likely not to be counted during surveys but because of the uncertainties involved in the surveys, the numbers are normally presented as minimum estimates of population size. It is on this basis that the most recent count totalling about 34,600 common seals is likely to indicate a total population of 50,000-60,000 seals.

Scotland During 2001, common seal distribution and abundance was estimated for Orkney and Shetland. This followed surveys in 2000 of the Inner and Outer Hebrides, Firth of Tay and the Inner Moray Firth. These surveys were mainly targeted at sites proposed as Special Areas for Conservation under the European Habitats Directive and were part-funded by Scottish Natural Heritage. A previous analysis of data from surveys of common seals along the north and west coasts of Scotland as far south as the southern tip of the Mull of Kintyre, and in Orkney, Shetland and the Hebrides indicated that there was an overall increase of about 3% per year in the number of animals counted at haul-out sites between 1988 and 1997. However, because of a high level of uncertainty, these changes could be caused by differences in the behaviour or distribution of seals between surveys, or in the environmental conditions when surveys were carried out. Comparing the surveys carried out in 2000 with those in 1996/97 suggested that there were more common seals in the west Highland and Strathclyde regions and a small decline in numbers in the Outer Hebrides and the Inner Moray Firth. Data from the surveys carried out by the University of Aberdeen suggest a continued decline in numbers in the Inner Moray Firth. The surveys conducted in 2001 indicated that common seal numbers had declined by 9% and 18.5% in Orkney and Shetland respectively since the previous survey in 1997. However, the counts in 2001 were within the range of the counts obtained from the previous three surveys carried out in these regions since 1989. Considering the uncertainties surrounding the counting methods, it is not possible to conclude from the data currently available that there has been a change in the size of the common seal population in Orkney and Shetland during the past 10 years.

The Wash and eastern England One complete aerial survey of common seals was carried out in Lincolnshire and Norfolk during August 2001. The count for The Wash (3194) was the highest ever recorded. It was 15% greater than the mean of the 2000 counts (2,778) and 5% greater than the higher of the 2000 counts (3,029). The average annual rate of increase in the number of seals counted in The Wash since 1989 is 6.3% (SE = 0.60%). This is significantly greater than the 3.5% (SE = 0.29%) average annual rate of increase between 1968 and 1988. Minimum estimate of the British common seal population Surveys of Orkney and Shetland were carried out in 2001. The most recent minimum estimate of the number of common seals in Scotland is 30,196 from surveys carried out in 1996, 1997, 2000 and 2001. The most recent minimum estimate of common seal numbers for the east coast of England is

6 SCOS 02/2 4,409. This comprises 4,274 seals in Lincolnshire and Norfolk in 2001 plus 135 seals in Northumberland, Cleveland, Essex and Kent between 1994 and 1997.

Possible effects of phocine distemper virus (PDV) The 2001 count of common seals in The Wash exceeded the 1988 pre-epidemic count by 5%. It has taken 13 years for the population to recover from the effects of the PDV epidemic. This is in contrast to populations on the east and south sides of the North Sea, which recovered more rapidly and were similar to or exceeded their pre-epidemic levels by 1996. During the previous outbreak of phocine distemper virus it is thought that about 10% of the common seal population in Scotland and up to 50% of common seals in England may have died. The current PDV outbreak has strong similarities in terms of its geographical origins and the chronology of infection to the outbreak in 1988 but there are detailed differences that could be important.

Counts by region are given in the Table below. These are minimum estimates of the British common seal population which is thought to be approximately 50,000-60,000 animals.

Region 1996-2001 Shetland 4,883 Orkney 7,752 Outer Hebrides 2,413 Highland (Nairn to Appin) 6,291 Strathclyde (Appin, Loch Linnhe to Loch Ryan) 7,909 Dumfries & Galloway (Loch Ryan to English Border 6 at Carlisle) Grampian (Montrose to Nairn) 126 Tayside (Newburgh to Montrose) 165 Fife (Kincardine Bridge to Newburgh) 611 Lothian (Torness Power Station to Kincardine Bridge) 40 Borders (Berwick upon Tweed to Torness Power 0 Station) TOTAL SCOTLAND 30,196

Blakney Point 772 The Wash 3,194 Donna Nook 233 Scroby Sands 75 Other east coast sites 135 South and west England (estimated) 20 TOTAL ENGLAND 4,429 TOTAL BRITAIN 34,625

7 SCOS 02/2

Questions from the Scottish Executive Environment and Rural Affairs Department

General How can population dynamics analysis models assist in estimating the potential impact of disease or population management measures on seal populations?

The response to the question is in three parts: an introduction of the modelling used to understand seal population dynamics and then separate responses to the two parts of this question.

Population modelling Population models use the underlying process of birth, death, immigration and emigration to generate the likely trajectory of seal populations. By fitting these models to data from population counts it is possible to gain an understanding of the processes driving change within a population. However, there can be considerable uncertainty surrounding the processes that affect birth and death rates, as well as those that affect immigration and emigration. This can lead to several plausible models, all of which are consistent with the available data. This suite of plausible models can then be used to forecast the possible range of future trends in the population.

Where British grey seals are concerned, uncertainty around these predictions arises from several sources:

(1) Estimates of birth rates for the population as a whole are based entirely on data collected in the 1970’s and early 1980’s. SMRU has no plans to use culling to obtain new data and there is no guarantee that the data from the 1980s apply to the current population. New methods involving the long-term observation of individuals at breeding colonies are being used to augment these older data.

(2) Death rates have been estimated from the age structure of seals shot in the 1970’s and these also may not be accurate in the present day. Again, new methods of mark-recapture are being developed and these could provide some information about death rates.

(3) There is very little information about rates of migration between individual seal colonies. However, new techniques to estimate these rates from genetic and mark-recapture experiments are being developed.

(4) We have relatively little information about how birth, death, immigration and emigration rates are likely to vary with the density of animals in the population.

On the positive side, there is greater certainty in the estimated number of pups born each year. The value of these estimates is enhanced by the long time-series (stretching back to the late 1960’s in many areas) and the fine spatial resolution of these estimates because they are specific to each breeding colony.

The approach being adopted to the modelling of the grey seal population by the Sea Mammal Research Unit (SMRU) is, therefore, to develop models that use birth/death and immigration/emigration rates that explain the greatest level of variation in the spatial and temporal distribution of the number of pups being born.

In collaboration with the Centre for Research into Ecological and Environmental Modelling at the University of St Andrews, SMRU is developing a fully spatially explicit model of British grey seal population dynamics that will be used to forecast trends in abundance.

8 SCOS 02/2 The situation for modelling the changes in common seal populations and the effects of management is considerably less developed than for grey seals. There is no equivalent base line of data about number of births each year. We currently have to use data for populations from outside the UK, although recent mark-recapture studies in the Moray Firth will provide information about birth and death rates. Therefore models for common seals with equivalent capabilities to those for grey seals are not feasible at present. More generalised models of seal population dynamics can provide very broad predictions of the effects of different management regimes.

Assessment of the impact of disease The situation is different in the case of disease epidemics, to which common seals appear to be more vulnerable than most other seal species. Predictive models of the way in which phocine distemper virus (PDV) spreads through a common seal population have been developed by the University of Cambridge in collaboration with SMRU. However, these models are sensitive to the characteristics of the disease. New strains of a disease like PDV could result in degradation of the predictive power of these types of models.

Nevertheless, the current models of PDV predict that the disease is unlikely to recur less than 10 year after a previous outbreak. The current outbreak has recurred 14 years after the previous outbreak. In future, it will be possible to build this type of disease process into management models for common seal populations.

Assessment of population management measures A good population model is an essential tool for assessing the effectiveness of management measures. A model that has been demonstrated to be realistic can be used to experimentally manipulate the population in advance of any management decisions being made. This will allow examination of the outcome of different management scenarios. Much of the uncertainty about possible outcomes of different management measures arises because of the uncertainties surrounding the data used to model seal populations. Nevertheless, it is still possible to examine management options using a precautionary approach. If the objective was to be precautionary about seal populations, this would entail running models for the worst case rather than the most likely case.

SMRU is developing a model that will allow this approach. This will permit SMRU to investigate the potential outcomes of different management approaches if or when they are proposed.

Grey Seals

Is there any evidence to suggest that the Scottish grey seal population is stabilising?

In the recent past, the UK Special Committee on Seals has noted that the rate of increase in pup production amongst grey seals has slowed in some regions. For example, in the Hebrides, pup production has not increased significantly since 1992. The pattern observed in the Hebrides could be being repeated in the Orkneys where pup production increased at 9.6% in the five years to 1996 and was 4.4% in the past 5 years. (Since number of births varies between years, it is advisable to consider percentage changes in pup production averaged over about 5 years). However, it should be noted that the overall trend in pup production for the total population is still upwards and it is still too early to say if the rate of increase in pup production in Orkney has changed significantly.

Pup production is used as an indicator of population size in grey seals. This assumes that changes in pup production are not compensated by changes in other factors like the juvenile survival rate. Changes in pup production can also take many years to filter through to equivalent changes in the population as a whole. If the increase in pup production has declined because of reduced birth rates then it could take more than 10 years before there is an equivalent change in the size of the entire population. This is because grey seals are very long-lived. However, if the rate of increase in pup production has declined because of reduced survival of adults then the change in pup production

9 SCOS 02/2 reflects a change in the rate of increase in the population as a whole. At present, there is no way of distinguishing between these possibilities but the former scenario, in which it may take 10 or more years for the population as a whole to follow the trend in the pup production, is believed to be the more likely.

Overall, the balance of evidence indicates that the grey seal population in Scotland will eventually stabilize, but it is not possible to say precisely when this will occur. It is likely that this will happen first in the Hebrides. The North Sea colonies as a whole are showing a gradual overall decline in the rate of increase in pup production and the fluctuations in pup production in Orkney in recent years could indicate that the population is approaching its limits

Is there any evidence that new seal breeding sites are developing and is such growth monitored?

There is some evidence of the development of new breeding sites for grey seals. For example, at Fast Castle (Berwickshire) a colony developed about 7 years ago and initially increased quickly (Appendix to Annex II, Table 4). Two additional colonies have formed in Orkney in recent years and the colony at Helmsdale, Sutherland, has increased through colonisation of previously unused beaches between Berriedale and Duncansby Head. All new colonies are included in the Sea Mammal Research Unit’s (SMRU) annual surveys of grey seal colonies. This is SMRU’s standard procedure when monitoring small breeding colonies.

Although SMRU does not have the capability to search all potential breeding colonies annually, SMRU monitors over forty existing small colonies within a three to five year cycle. Many of these colonies are surveyed in alternate years. In addition, islands and coasts that appear to provide suitable breeding habitat are also surveyed and the annual grey seal surveys allow inspection of large areas of coastline where new colonies could have formed. SMRU also incorporate information passed on by other observers, particularly regional staff of Scottish Natural Heritage. Grey seals can also give birth at unusual sites from time-to-time. Nevertheless, the total number involved forms a very small proportion (likely to be <1%) of the total pups born at other sites and, as such, does not presently constitute a significant proportion of the UK population.

If so, how is this reflected or how can this be reflected in population estimates?

The total population is estimated from the pup production. The proportion of pup production not accounted for at the annually monitored sites is estimated to be less than 15% of the total. The information from these sites is obtained on a less formal basis but is, nevertheless, included in the overall population estimates. The information from these other sites is tabulated annually in the Advice (Annex II, Table 2). The region requiring the most urgent attention is Shetland, which was previously surveyed in its entirety in 1977, with a partial survey in 1994 by E. Brown, under contract to Scottish Natural Heritage.

How can the latest information on habitat use by grey seals outwith the breeding season be used to further our understanding of their interactions with fisheries?

Over the past 10 years, SMRU has undertaken studies to examine the distribution of grey seals outside the breeding season. This has involved tracking individuals at sea and developing statistical methods to extrapolate from this sample to the whole population. The latest summary information is provided in Figure 1. Further details are provided in Annex V, Information Paper 3. This shows clear concentrations of foraging effort in grey seals. It is not possible to say what impact this distribution might have on marine fish stocks, but two important features are worth noting: (1) most of the foraging by grey seals occurs offshore (>10 miles from land) and is not associated with regions immediately surrounding the locations where they are observed ashore and (2) these areas are not necessarily those used most intensively by fisheries.

10 SCOS 02/2

SMRU will continue to improve this estimated distribution by accounting for the different behaviour of males and females, and juvenile adult seals, and by considering changes depending upon time of year.

Figure 1. The estimated distribution of grey seals around the UK coast at times of year other than the breeding season and when not hauled out on land. The estimate is based upon the distribution of 108 grey seals tracked using satellite tags.

Common Seals

How could current estimates of common seal numbers and in particular the identification of population trends be improved?

Recent estimates of common seal populations have been made using aerial surveys which provide a synopsis of common seal abundance in large regions of coastline (each region is about one-fifth of the coastline of Scotland). In the past, SMRU has tried to survey at least one of these regions each year. Only in The Wash, the Moray Firth and the Firth of Tay is there an annual count of common seals. It has not been possible for SMRU to maintain its schedule of surveys in recent years, although specific regions have been surveyed more regularly under contract to Scottish Natural Heritage

These surveys provide information about the minimum number of seals within a region at the height of the moult (August). The moulting period is chosen because this is the time of year when the number of seals on land is greatest and most consistent. The method can only count seals while they are on land and, in order to translate this into a total population size, it is then necessary to adjust the estimate to account for those seals that were in the water at the time of the survey. This requires a detailed knowledge of the movements of a sample of individual common seals and this information is only available for a few of the regions surveyed and these mainly include sites where seals haul out on sand banks. This type of information is expensive to obtain, although new technology is helping to increase the practicality of the approach. SMRU has plans to undertake tracking of common seals in the Hebrides and is currently carrying out tracking studies on the east coast of Scotland.

To maximise the effectiveness of surveying, it will probably be necessary also to examine common seal numbers at specific sites more frequently and over longer time periods within a year than is possible using current methods. These could be described as trend sites. The degree to which these trend sites are representative of the whole population could be assessed from the synoptic aerial surveys. However, more research is required to assess the number of trend sites that will be required to provide reliable information.

11 SCOS 02/2

Is it possible to estimate the potential impact of the latest Phocine Distemper Virus outbreak on Scottish Common seal populations?

The best indicator of the potential impact of Phocine Distemper Virus (PDV) on common seals in Scotland is the pattern of infection and death that were observed during the previous outbreak in 1988. The source and subsequent radiation of the current outbreak is following a similar pattern to that of 1988, although the disease appears to be taking longer to spread between local populations. During the 1988 outbreak common seals in Scotland were affected much less than anywhere else in Europe. Large numbers of dead seals were reported from the Moray Firth , Orkney, the west coast and the Clyde. However, when these populations were resurveyed in 1989, the only detectable decline in numbers (of around 10%) was on the east coast.

We do not know why common seals in Scotland were affected to a lesser extent than those in the southern North Sea. However, serological testing suggested that over 80% of Scottish seals were exposed to the virus suggesting that they were more resistant to the disease than seals in the southern North Sea and Baltic. Recent serology studies indicate that most common seals in the Scottish population are once again susceptible to the PDV virus.

Seal Diet

What progress is being made in the latest study of grey seal diet?

The Sea Mammal Research Unit is currently carrying out a study of grey seal diet in the North Sea, Orkney, Shetland and the Hebrides. This is funded by DEFRA, SEERAD and SNH and its objectives are to repeat a study of diet carried out by SMRU in 1985. A wider area will be covered in 2002 than in 1985. This is a 3-year project which is still in its first year. Scats have been collected in the Hebrides, Orkney, Shetland and along the North Sea coast beginning in January 2002. These samples are currently being analysed and further collections will take place across all these regions to ensure a representative set of samples is obtained throughout the year.

The Impact of Seals on Salmonids

Is it possible to establish whether individual seals specialise in salmonids? If so, is it possible to estimate their impact on overall salmonid mortality?

It may be possible to establish if individual seals specialise in salmonids using methods that determine the relative abundance of fatty acids (i.e. building blocks of fat) in different fish species or species groups in the diet. This method requires knowledge of the fatty acids incorporated into the tissues of seals not only from salmon but also from all other potential prey species, including sea trout which may be particularly difficult to distinguish from salmon prey. In collaboration with the Scottish Fisherman’s Federation, SMRU is planning a project that, if funded, will provide a library of chemical profiles for each prey species that takes account of variations among regions and at different times of year. It will then be possible to compare profiles from individual seals with this library.

Estimating the contribution of seal predation to overall salmon mortality requires knowledge of all the other forms of mortality experienced by salmon. At present in Scotland there is a relatively large population of seals and small populations of salmon. Therefore, even if only a small proportion of the seal population eat adult salmon occasionally, they could have a substantial effect upon salmon populations. This has two consequences: salmon predation is likely to be a rare event when viewed in terms of the average seal, and it is difficult to obtain reliable estimates of seal predation on salmon.

12 SCOS 02/2 It is likely that the greatest effects of seals on salmon occur in the entrances to the main rivers and this is the region in which we suggest that future research should be concentrated. There may be a case for carrying out experimental studies of seals that specifically occupy river systems to establish the extent to which they predate salmon, whether there are methods that could be used to create seal exclusion zones at some sites and whether the removal of “rogue” seals would be an effective strategy to reduce predation of salmon. However, studies of grey seals in the Baltic showed that removal of “rogue” seals had no effect on the apparent damage caused by seals to a coastal salmon fishery

What are the principle areas of uncertainty concerning seal/salmonid interactions and which of these areas might be tackled by focussed research?

This question can be answered on two levels because seal/salmonid interactions can be viewed as (1) an effect of seals on the population of salmon as a whole and (2) the effect of seals on salmon fisheries.

Considering the effect upon salmon populations as a whole, seals are only one of several different sources of mortality for salmon that could influence the size of a population and, therefore, the level of a fishery. There are large parts of the salmon life-history, from the production of smolts to the return of adults to rivers, that could be influenced by both environmental change/variability and by predation from species other than seals. At this stage, it is impossible to determine the principal uncertainties, although this question could merit specific theoretical studies to examine the sensitivity of salmon life-histories to predation at different stages. Seals are known to contribute to the mortality of salmon at various stages in their life cycle but the importance of seal predation for salmon populations as a whole is unknown.

A principal uncertainty involved in this assessment is whether seals are involved in predation of salmon in the open sea. Since salmon are large highly mobile fish, it is considered unlikely that seals will have the capability regularly to capture salmon in open water. Resolution of this question could come from focussed research on the diet of individual seals using fatty acids (as described above) and using state-of-the-art technology to study seal foraging behaviour Considering the direct effects upon salmon fisheries, recent analyses of seal interactions with fisheries in the northeast Atlantic region suggest that seals are opportunists and will learn to use fishing gear to help capture highly mobile prey like salmon if the design of the fishing gear has not taken deterrence of seal predation into account. Therefore, focussed research on the design of fishing gear to reduce the capacity of seals to capture or damage salmon could be fruitful. Some new forms of net design used in Sweden are proving to be successful in this regard.

Non-lethal Methods of Population Control

Have there been any recent developments in relation to non-lethal methods of population control which could be effectively applied to Scottish seal populations?

SMRU is continuing to monitor developments in this field. The method of non-lethal population control that is most often proposed for Scottish seal populations is the application of immuno- contraception that renders females sterile for a number of years.

Recent applications of the method to the control of white-tailed deer and horse populations in the United States have apparently proved to be successful, at least in terms of the efficacy of the treatment to induce a degree of sterility. The long-term success as a measure of controlling larger populations remains to be tested.

13 SCOS 02/2 There are a number of problems with applying this method to the control of grey seal populations. These include the degree to which it is possible to implement such a system of control across a sufficiently large section of the population, the costs involved and the extent to which the objective of reducing the impact of seals on fisheries might be achieved. Nevertheless, a small-scale trial to assess the biological side-effects, efficacy and the economics of such a system of population management in grey seals might be useful.

14 SCOS 02/2 NERC Special Committee on Seals

Terms of Reference 1. To undertake, on behalf of Council, the provision of scientific advice to the Scottish Executive and the Home Office on questions relating to the status of grey and common seals in British waters and to their management, as required under the Conservation of Seals Act 1970. 2. To comment on SMRU’s core strategic research programme and other commissioned research, and to provide a wider perspective on scientific issues of importance, with respect to the provision of advice under Term of Reference 1.

3. To report to Council through the NERC Chief Executive.

Current membership Professor JR Beddington (Chairman), Imperial College, London; Dr WD Bowen, Bedford Institute of Oceanography, Halifax, Nova Scotia, Canada; Professor IL Boyd, University of St Andrews; Professor JH Lawton, Chief Executive, NERC, Swindon; Dr A McLay, FRS Marine Laboratory, Aberdeen; Dr EJ Millner-Gulland, Imperial College, London; Dr P Reijnders, Institute for Forestry and Nature Research, Texel, The Netherlands; Dr JK Pinnegar, CEFAS Fisheries Laboratory, Lowestoft; Professor W Sutherland, University of East Anglia; Dr PM Thompson, University of Aberdeen; Professor F Trillmich, University of Bielefeld, Germany; Dr F Carse (Secretary), NERC, Swindon.

15 SCOS 02/2 SCOS 02/02 ANNEX I The Status of Grey and Common Seals in the North Atlantic Region

Table 1 Sizes of grey seal populations Region Population Year when latest Type of data (see Population status size information was key1) obtained Mainland Scotland & 10,800 1998-2001 1 Shetland Outer Hebrides 38,100 2001 2 Pup production stable, total population probably still increasing Inner Hebrides 9,100 2001 2 Pup production stable, total population probably still increasing Orkney 54,200 2001 2 Increasing Scottish North Sea 7,000 2001 2 Increasing coast Scotland 119,200

English North Sea 5,900 2000 2 Increasing coast Southwest 4,600 1999 1 Stable (England/Wales) England & Wales 10,500

Total (UK) 129,700

Ireland 2,000 1997-99 1 Unknown Norway 3,000-3,500 1986 1 Unknown Germany 71 1991 1 Increasing The Netherlands 500 2000 1 Increasing Baltic 12,053 2000 1 Increasing Iceland 11,600 1987 1 Faroes 4,000 1966 1 Barents Sea 3,400 1990 1 Europe (excluding 36,600 UK)

Canada 173,500 1998 2 Increasing

Total 339,800

1 1 – Estimates based upon occasional pup counts 2 - Estimates based upon systematic annual pup counts using aerial survey

16 SCOS 02/2 Table 2 Sizes and status of European populations of common seals. Region Population size1 Years when latest Status information was obtained Outer Hebrides 2,400 1996-2000 Possible increasing Scottish W coast 14,200 1996-2000 Possible increasing Scottish E coast 900 1996-97 Stable Shetland 4,900 1996-2001 Possible decrease Orkney 7,800 1996-2001 Possible decrease Scotland 30,200

England (E & S coast) 4,400 2001 Increasing

Northern Ireland 400 1997 Decrease since 1970s

UK (Total) 35,000

Ireland 900 1978 Wadden Sea (Germany) 11,500 2000 Increasing Wadden Sea 3,300 2000 Increasing (Netherlands) Wadden Sea (Denmark) 2,100 2000 Increasing Limfjorden (Denmark) 1,000, 495 1998-2000 Decrease since 1998 Kattegat/Skagerrak 9,752 2000 Increasing West Baltic 315 1998 Small increase Kalmarsund (East 270 1998 Increasing Baltic) Norway S of 62ºN 1,200 1996-98 Unknown Norway N of 62ºN 2,600 1994 Unknown Iceland 19,000 ? Unknown Barents Sea 660 ? Unknown Europe (excluding 52,600 UK)

Total 87,500

1 – many of these estimates represent counts of seals. They should be considered as minimum estimates of total population size.

17 SCOS 02/2

SCOS 02/2 ANNEX II Pup Production in the British Grey Seal Population

Callan Duck

Sea Mammal Research Unity, Gatty Marine Laboratory, University of St Andrews

1. Surveys conducted in 2001 Each year SMRU conducts aerial surveys of the major grey seal breeding colonies in Britain to determine the number of pups born. In addition, new sites where grey seal pups have been reported or which appear to be suitable for colonisation are visited regularly. During 2001, between four and six surveys were flown over all the major sites in the Hebrides, Orkney and in the Firth of Forth. Ground counts of the numbers of pups born at the Farne Islands were made by National Trust staff. Similar counts at Donna Nook on the Humber Estuary were made by members of the Lincolnshire Wildlife Trust. Locations of the main British grey seal breeding sites are shown in Figure 1.

2. Estimated pup production The number of pups born (pup production) at regularly surveyed colonies is estimated each year from counts from aerial survey photographs using a model of the birth process and the development of pups. The method used to obtain the estimates for this year’s advice was similar to that used for the past several years. Total pup production in 2001 at all annually surveyed sites is estimated to be 36,920 which is almost identical to the estimate for 2000 (36,915 pups). Estimates of the total pup production from all major breeding sites in England and Scotland (except Loch Eriboll, Helmsdale and Shetland) between 1984 and 2001 are shown in Figure 2. Pup production estimates for the main island groups (the Inner Hebrides, the Outer Hebrides and Orkney) are shown in Figure 3a and for the North Sea sites in Figure 3b. The time series of data for island groups are given in Table 1. For colonies not surveyed by air, pup numbers are counted directly on the ground either annually (Farne Islands and Donna Nook,) or less frequently (South Ronaldsay, SW England, Wales and Shetland).

3. Trends in pup production Between 1984 and 1996 estimates of the total number of pups born at regularly surveyed colonies have increased year on year. In 1997, estimated pup production fell for the first time but recovered again in 1998 in line with the previously observed upward trend. However, there was a second temporary decline in 1999 followed by a recovery in 2000. Pup production remained nearly static between 2000 and 2001. The differences between 2000 and 2001 are shown in Text-table 1. This shows that, while the percentage change between 2000 and 2001 varied from –10% at the Isle of May to +10% in Orkney, the overall pup production for the population as a whole was static between 2000 and 2001. There appears to be greater fluctuation in pup production within regions between years. For example, the change between 1999 and 2000 in the Outer Hebrides was +15% whereas the change in the past year for the same region was –8%. A similar pattern was observed for the Inner Hebrides, which had a similar magnitude of fluctuation to that of the Outer Hebrides suggesting that the fluctuations in the pup production in these two regions could be caused by the same process. Until 2001, fluctuations in

18 SCOS 02/2 Orkney also followed a similar pattern to those in the Outer Hebrides. However, in 2001 pup production in the Outer Hebrides declined while it increased in Orkney. The mean annual percentage change in pup production for each region is shown in Text-table 1. This shows that, apart form Donna Nook, which is the smallest of all the regional groups considered here with only 634 pups born in 2001, the mean annual change ranged from –2% to +4% and the annual change for the population as a whole was +2.8%

Text-table 1. The percentage change in grey seal pup production at annually surveyed colonies, between 2000 and 2001 with the mean annual change between 1997 and 2001.

Location Change 2000-2001 Mean Change 1997-2001 Inner Hebrides -8.8% -0.5% Outer Hebrides -8.0% +1.4% Orkney +9.6% +4.4% Isle of May + Fast Castle -10.4% +3.3% Farne Islands +6.5% -1.7% Donna Nook +2.6% +14.5% Total 0.0% +2.8%

The results from 2001 support the trends observed in recent years. First, pup production in the Hebrides has remained unchanged since 1992. Second, based upon the most recent count, there is now reduced evidence to support the suggestion of a slowing in the rate of increase in pup production in Orkney. Third, the North Sea colonies as a whole are showing a gradual overall decline in the rate of increase in pup production even though there has been a rapid increase at the small colony at Donna Nook on the Lincolnshire coast (Text-table 2). Overall, the trend in the rate of increase suggests a gradual decline has been taking place during the past 10 years. In the late 1980s, pup production was increasing at over 6% per annum and this had declined to 3% in the past 5 years.

Pup production fluctuates between years, but in the last 5 years the fluctuations have been larger than previously (Figure 2). This is also reflected in the annual rate of change between years (Annex IV, Figure 4). It is difficult to be sure what causes these changes but they could indicate that the population is nearing its limits of size. To even out these fluctuations the average percentage rate of annual change in the pup production by region is shown in Text-table 2 for the past 5 years and this probably provides the best indication of the current trend in pup production.

4. Pup production model assumptions The model used to estimate pup production from aerial survey counts of whitecoat and moulted pups assumes that the parameters defining the distribution of birth dates are variable from site to site and year to year, but that those defining the time to moult and time to leave the colony remain constant. The pup production estimate is sensitive to the value used for the latter parameter and hence there is an argument for allowing this parameter to vary between colonies. In the past versions of this advice, we have considered the effect of allowing the time-to-leave parameter to vary. However, although the pup production trajectory is slightly lower using the method with variable time-to-leave, the variations in pup production are consistent amongst the two methods. Since we are in the process of developing a new method for estimating pup production from production curves we will not present data using the method involving a variable time-to-leave.

19 SCOS 02/2 This is consistent with the Advice provided in previous years.

Text-table 2. Pup production estimates for the main colonies surveyed in 2001. The annual changes over successive 5-year periods are also shown. These annual changes represent the exponential rate of change in the pup production. The total for the North Sea represents the estimates for the Isle of May, Fast Castle, Farne Islands and Donna Nook combined.

Location 2001 pup Annual change in pup production production 1987- 1992- 1997- 1991 1996 2001 Inner 2,938 +4.2% +2.5% -0.5% Hebrides Outer 12,325 +5.1% +2.2% +1.4% Hebrides Orkney 17,523 +7.4% +9.6% +4.4% Isle of May + 2,253 +14.0% +3.9% +3.3% Fast Castle Farne Islands 1,247 +2.1% +1.7% -1.7% Donna Nook 634 +39.0% +14.6% +14.5% Total (North 4,134 +9.6% +4.0% +3.0% Sea) Total 36,920 +6.2% +5.2% +2.8%

6. Confidence limits Ninety-five percent confidence limits on the pup production estimates at each site are within 14% of the point estimate. The exact limits depend on a number of factors including the number of surveys flown in a particular year. Confidence limits can be seen in Figures 3a (Orkney, Inner and Outer Hebrides) and 3b (Isle of May and Fast Castle only).

7. Pup production at sites surveyed less frequently The total population associated with breeding sites not surveyed regularly has been calculated using the ratio of total population to pup production for the main areas. Less than 15% of all pups are born at these sites each year. Confidence limits cannot be calculated for these estimates because they are obtained by simple extrapolation of single counts. The resulting figures are given in Text-table 3.

20 SCOS 02/2

Text-table 3. Pup production estimates for breeding sites not surveyed regularly

Region Date and location of last Pup production (to survey nearest 100) Mainland Scotland Helmsdale (including & South Berriedale) 2001 1,800 Ronaldsay Loch Eriboll 2000 South Ronaldsay 1998 Shetland 1977 1,000 Southwest Britain Southwest England 1973 1,500 Wales 1994

Text-table 3 shows Scottish breeding sites which are either not surveyed annually or have recently been included in the survey programme. These and other potential breeding sites are checked when flying time, flying conditions and additional circumstances permit. Accumulated data from sites that are surveyed on an ad hoc basis are given in Table 2. Taking all these additional sites into account, about 5,000 pups are likely to be born at sites that are not part of the main annual survey.

21 SCOS 02/2 Table 1. Estimates of pup production for the North Sea, Orkney, Outer Hebrides and Inner Hebrides, 1960-1999.

YEAR North Sea Orkney Outer Inner Hebrides Hebrides 1960 1020 2048 1961 1141 1846 3142 1962 1118 1963 1259 1964 1439 2048 1965 1404 2191 1966 1728 2287 3311 1967 1779 2390 3265 1968 1800 2570 3421 1969 1919 2316 1970 2002 2535 5070 1971 2042 2766 1972 1617 4933 1973 1678 2581 1974 1668 2700 6173 1975 1617 2679 6946 1976 1426 3247 7147 1977 1243 3364 1978 1162 3778 6243 1979 1620 3971 6670 1980 1617 4476 8026

22 SCOS 02/2 Table 1 continued.

YEAR North Sea Orkney Outer Inner Hebrides Hebrides 1981 1531 5064 8086 1982 1637 5241 7763 1983 1238 1984 1325 4741 7594 1332 1985 1711 5199 8165 1190 1986 1834 5796 8455 1711 1987 1867 6389 8777 2002 1988 1474 5948 8689 1960 1989 1922 6773 9275 1956 1990 2278 6982 9801 2032 1991 2375 8412 10617 2411 1992 2437 9608 12215 2816 1993 2710 10790 11915 2923 1994 2652 11593 12054 2719 1995 2757 12412 12713 3050 1996 2938 14195 13176 3117 1997 3698 14051 11946 3076 1998 3989 16231 12373 3087 1999 3380 15253 11683 2787 2000 4303 15993 13396 3223 2001 4134 17523 12325 2938

23 SCOS 02/2 Table 2. Scottish grey seal breeding sites that are not surveyed annually and/or have recently been included in the survey programme.

Location Survey method Last surveyed, Number of pups frequency Inner Colonsay/Oronsay mainland SMRU visual 2001, annual 94 Hebrides Loch Tarbert, Jura SMRU visual 1998, every 3-4 years None seen West coast Islay SMRU visual 1998, every 3-4 years None seen Ross of Mull, south coast SMRU visual 1998, infrequent None seen Treshnish small islands, incl. SMRU photo & 1999, annual ~20 in total Dutchman’s Cap visual Staffa SMRU visual 1998, every other year ~5 Little Colonsay, by Ulva SMRU visual 1998, every 3-4 years 6 Meisgeir, Mull SMRU visual 1998, every 3-4 years 1 Craig Inish, Tiree SMRU photo 1998, every 2-3 years 2 Cairns of Coll SMRU photo 1998, every 2-3 years 13 Muck SMRU photo 1998, every other year 12 Rum SNH ground 2001, annual 10-15 Canna SMRU photo 1998, every other year 34 Rona SMRU visual 1989, infrequent None seen Ascrib Islands, Skye SMRU photo 1998, every other year 32 Heisgeir, Dubh Artach, SMRU visual 1995, every other year None Skerryvore 1989, infrequent None Outer Barra Islands SMRU visual 2001, annual? 102 Hebrides Fiaray & Berneray Sound of Harris islands SMRU photo 1999, every 2-3 years 317 St Kilda Warden’s reports Infrequent Few pups are born Shiants SMRU visual 1998, every other year None Flannans SMRU visual 1994, every 2-3 years None Bernera, Lewis SMRU visual 1991, infrequent None seen Summer Isles SMRU visual 1989, infrequent None seen Faraid Head SMRU visual 1989, infrequent None seen Eilean Hoan, Loch Eriboll SMRU visual 1998, annual None Rabbit Island, Tongue SMRU visual 1998, every other year None seen Orkney Sule Skerry SMRU photo 1998, 1999, 2000, 2001 15, 7, 7, 10 Sanday, Point of Spurness SMRU photo 1999, every 2-3 years 62 Sanday, east and north SMRU visual 1994, every 2-3 years None seen Papa Stronsay SMRU visual 1993, every 3-4 years None seen Holm of Papa, Westray SMRU visual 1993, every 3-4 years None seen North Ronaldsay SMRU visual 1994, every 2-3 years None seen Eday mainland SMRU photo 2000, first 8 Calf of Flotta SMRU photo 2000, annual 250 South Fara, Cava & Rysa SMRU photo 2000, annual 155 Others Firth of Forth islands & Anecdotal Infrequent <10 Inchcolm SMRU photo 1997 4

24 SCOS 02/2

Figure 1 Map of the UK showing the locations of grey seal breeding colonies

25 SCOS 02/2 Figure 2. Total estimated pup production for all major breeding colonies in Scotland and England (excluding Loch Eriboll, Helmsdale and Shetland) from 1984 to 2001.

Grey seal pup production at annually surveyed sites

40000 38000 36000 34000 32000 30000 28000 26000 24000 22000 20000 18000 16000 14000 Number of pups 12000 10000 8000 6000 4000 2000 0 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 Year

26 SCOS 02/2 Figure 3 Trends in pup production at the major grey seal breeding areas since 1984. Production values are shown with their upper and lower 95% confidence limits where these are available. These limits assume that the various pup development parameters which are involved in the estimation procedure remain constant from year to year. Although they therefore underestimate the total variability in the estimate, they are useful for comparison of the precision of the estimates in different years. Note that the scale of these two figures differs by an order of magnitude. (a) Outer Hebrides, Orkney and Inner Hebrides

Grey seal pup production at the main island groups

20000 Outer Hebrides Orkney 18000 Inner Hebrides 16000

14000

12000

10000

8000 Number of pups 6000

4000

2000

0 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 Year

(b) Isle of May, Farne Islands and Donna Nook

Grey seal pup production at North Sea sites

3000 Farnes 2800 Isle of May + Fast Castle 2600 Donna Nook 2400 2200 2000 1800 1600 1400 1200 1000 Number of pups 800 600 400 200 0 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 Year

27 SCOS 02/2 SCOS 02/2 ANNEX II Appendix

Estimates of pup production

Table 1. Pup production estimates for islands in the Inner Hebrides group

YEAR Gunna Northern Fladda Sgeir a’ Lunga Soa Eilean Eilean Nave TOTAL Treshnish Chaisteil & nan Ron nan Eoin Island Eirionnach 1984 206 87 169 136 226 63 180 190 75 1332 1985 192 84 109 113 136 63 158 269 66 1190 1986 263 114 149 119 204 111 302 305 144 1711 1987 361 115 194 147 234 102 420 297 132 2002 1988 332 130 231 170 246 102 389 225 135 1960 1989 347 131 234 187 277 101 308 167 204 1956 1990 342 146 183 162 221 107 392 265 214 2032 1991 475 125 288 174 271 97 409 377 195 2411 1992 527 203 347 153 341 98 453 438 256 2816 1993 514 211 324 186 385 91 464 458 290 2923 1994 580 145 280 148 356 96 349 456 309 2719 1995 541 181 368 182 429 116 454 440 339 3050 1996 583 181 351 186 414 92 558 431 321 3117 1997 589 158 365 177 448 81 562 414 282 3076 1998 638 168 315 166 427 63 490 430 390 3087 1999 522 158 319 173 352 57 481 392 333 2787 2000 625 177 355 167 392 82 617 406 402 3223 2001 615 171 295 141 350 74 515 405 372 2938

28 SCOS 02/2 Table 2. Pup production estimates for islands in the Outer Hebrides group

Shillay YEAR Gasker Coppay (Sound of Haskier Causamul Deasker Shivinish Ceann Iar Ceann Ear Shillay Stockay Monachs Others North TOTAL Harris) (Monachs) (Monachs) (Monachs) (Monachs) (Monachs) total Rona 1960 1961 847 62 120 81 67 13 0 0 1949 3142 1962 1963 1964 1965 1966 1084 230 120 96 242 0 0 38 0 1499 3311 1967 1084 153 80 96 161 0 0 114 0 1574 3265 1968 1084 115 161 96 161 0 0 152 0 1650 3421 1969 1970 1129 324 714 130 103 41 0 0 84 60 460 605 0 2023 5070 1971 1972 1141 316 605 167 271 67 0 0 274 49 730 1054 0 1309 4933 1973 1974 1756 286 692 176 224 83 0 49 459 44 754 1307 0 1647 6173 1975 1538 367 631 212 202 51 0 141 690 217 932 1982 0 1961 6946 1976 1813 394 553 278 217 57 0 111 628 152 1053 1946 0 1886 7147 1977 1978 1101 321 508 320 172 51 0 560 371 205 626 1764 0 2002 6243 1979 992 377 546 269 159 80 0 672 810 164 826 2474 0 1770 6670 1980 1345 462 794 351 163 31 0 1077 880 242 647 2848 162 1867 8026

29 SCOS 02/2

Table 2 (continued). Pup production estimates for islands in the Outer Hebrides group

Shillay YEAR Gasker Coppay (Sound of Haskier Causamul Deasker Shivinish Ceann Iar Ceann Ear Shillay Stockay Monachs Others North Rona TOTAL Harris) (Monachs) (Monachs) (Monachs) (Monachs) (Monachs) total 1981 1255 423 1016 278 178 68 0 1279 486 331 847 2944 136 1785 8086 1982 1443 634 219 322 260 110 0 1329 557 199 712 2798 85 1888 7763 1983 1984 1120 389 386 277 143 0 83 2175 616 209 555 3638 0 1641 7594 1985 1303 408 335 254 168 0 261 2365 748 193 641 4208 0 1489 8165 1986 1258 378 356 225 108 0 283 2931 822 222 572 4830 0 1300 8455 1987 1337 393 365 224 131 0 353 3227 666 223 670 5139 0 1188 8777 1988 1205 354 372 195 122 0 429 3733 418 189 579 5348 0 1093 8689 1989 1294 383 348 176 73 0 512 4041 518 212 535 5818 0 1183 9275 1990 1398 396 321 146 115 0 574 4554 510 174 457 6269 0 1156 9801 1991 1406 440 334 159 94 0 582 5098 543 181 494 6898 0 1286 10617 1992 1527 427 514 179 91 0 576 5852 716 204 599 7947 0 1530 12215 1993 1525 366 431 150 107 0 640 5498 1037 192 524 7891 0 1445 11915 1994 1432 394 491 123 86 0 640 5956 921 196 522 8235 0 1293 12054 1995 1389 392 570 120 55 0 856 6332 977 200 480 8845 0 1342 12713 1996 1508 391 574 133 64 0 721 6648 1254 157 445 9225 0 1281 13176 1997 1301 303 470 79 67 0 795 5660 1656 76 458 8645 0 1081 11946 1998 1444 307 552 90 64 0 865 5711 1649 70 422 8717 0 1199 12373 1999 1247 224 508 66 45 0 739 5637 1514 69 464 8423 0 1170 11683 2000 1388 212 519 80 48 0 834 6439 2199 129 443 10044 0 1105 13396 2001 1251 206 492 79 35 0 771 5724 2110 138 473 9216 0 1046 12325

30 SCOS 02/2

Table 3. Pup production estimates for islands in the Orkney group

Muckle Little Little Holm of Point of Linga- Holm Fara- Faray Rusk- Wart- Sweyn- Grass- Swona Pent- Aus- Switha Stroma Calf of Copin- Stron- YEAR Green- Green- Linga Spur- Spur- holm of holm holm holm holm & holm land kerry Eday say say TOTAL holm holm ness ness Huip Gairsay Skerry 1960 734 190 239 90 0 0 0 441 0 208 41 0 0 2 98 0 0 0 0 0 0 2048 1961 537 290 251 124 0 0 0 300 0 256 33 0 0 2 48 0 0 0 0 0 0 1846 1962 ...... 1963 ...... 1964 934 469 154 25 0 0 0 22 117 208 16 55 3 14 24 0 0 0 0 0 0 2048 1965 671 366 279 138 0 0 0 113 151 247 29 21 66 19 85 0 0 0 0 0 0 2191 1966 688 454 344 138 0 0 0 270 154 87 8 59 18 14 48 0 0 0 0 0 0 2287 1967 600 445 395 98 0 0 0 270 165 252 8 111 0 6 36 0 0 0 0 0 0 2390 1968 650 310 399 278 0 13 0 257 258 195 8 81 36 27 52 0 0 0 0 0 0 2570 1969 567 298 576 189 8 28 0 214 28 208 4 77 59 35 20 0 0 0 0 0 0 2316 1970 747 318 519 135 45 42 22 171 95 223 4 13 66 43 85 0 0 0 0 0 0 2535 1971 588 351 708 158 49 137 30 320 88 103 16 70 40 67 36 0 0 0 0 0 0 2766 1972 ...... 1973 503 207 519 233 66 177 88 351 35 15 12 86 92 51 52 87 0 0 0 0 0 2581 1974 525 190 479 146 21 61 137 500 72 132 0 134 69 71 73 84 0 0 0 0 0 2700 1975 483 230 483 271 49 39 117 477 65 63 4 111 21 59 48 152 0 0 0 0 0 2679 1976 605 175 648 328 53 68 68 398 85 60 4 198 21 92 65 375 0 0 0 0 0 3247 1977 679 210 684 305 78 50 130 477 58 111 4 194 21 92 65 199 0 0 0 0 0 3364 1978 333 210 800 471 136 79 192 700 58 219 4 149 36 104 57 134 0 90 0 0 0 3778 1979 546 294 344 430 127 144 368 672 92 280 4 142 69 92 65 145 0 152 0 0 0 3971 1980 496 166 676 415 107 315 275 817 165 336 0 167 74 108 81 97 0 174 0 0 0 4476 1981 442 199 860 449 45 293 510 712 202 319 4 108 92 225 125 249 0 223 0 0 0 5064

31 SCOS 02/2

Table 3 (continued). Pup production estimates for islands in the Orkney group

YEAR Muckle Little Little Holm of Point of Linga- Holm Fara- Faray Rusk- Wart- Sweyn- Grass- Swona Pent- Aus- Switha Stroma Calf Copin- Stron- TOTAL Green- Green Linga Spur- Spur- holm of holm holm holm holm & holm land kerry of say say holm -holm ness ness Huip Gairsay Skerry Eday 1982 454 87 716 665 29 326 521 817 146 295 4 104 103 148 147 294 153 227 0 0 0 5241 1983 ...... 1984 517 127 601 518 0 303 368 834 376 335 0 111 79 85 70 219 119 79 0 0 0 4741 1985 483 191 568 643 0 342 245 796 526 315 0 115 60 260 82 261 151 161 0 0 0 5199 1986 637 227 602 533 0 390 358 752 811 345 0 145 81 191 70 278 157 219 0 0 0 5796 1987 592 231 678 570 0 502 548 837 910 261 0 109 83 327 90 216 153 282 0 0 0 6389 1988 393 181 590 424 0 569 557 833 921 247 0 135 66 336 62 222 167 245 0 0 0 5948 1989 426 191 574 426 0 696 638 760 1452 232 0 164 48 314 62 279 207 304 0 0 0 6773 1990 334 201 625 341 0 807 731 970 1313 179 0 195 49 351 79 252 206 349 0 0 0 6982 1991 459 186 728 388 0 1144 880 976 1602 192 0 214 70 514 96 277 272 414 0 0 0 8412 1992 507 222 845 462 0 1186 1052 1304 1845 204 0 223 56 585 51 206 304 556 0 0 0 9608 1993 601 241 830 385 0 1249 1221 1325 1781 218 0 292 88 604 86 166 324 595 270 514 0 10790 1994 642 262 786 348 0 1527 1294 1238 1909 220 0 272 69 674 65 161 331 508 346 795 146 11593 1995 728 300 795 420 0 2128 887 1387 2136 251 0 461 32 578 71 125 442 339 274 940 118 12412 1996 770 289 834 416 0 2255 1349 1464 1935 243 0 518 64 829 79 123 370 583 399 1480 195 14195 1997 786 332 771 387 0 2294 1071 1464 2024 215 0 336 46 870 66 131 347 638 587 1455 231 14051 1998 883 442 842 429 0 2583 1323 1675 2166 272 0 405 61 1032 69 123 430 784 499 1914 299 16231 1999 790 438 632 449 0 2390 1240 1399 2152 220 0 505 40 957 62 44 449 686 567 1962 271 15253 2000 898 367 704 419 0 2890 1347 1293 2061 191 0 482 22 1005 60 54 474 826 456 2082 362 15993 2001 1000 427 723 482 3156 1402 1291 2168 239 563 26 1077 55 58 441 1091 556 2540 300 17523

32 SCOS 02/2

Table 4. Pup production estimates for other sites routinely monitored.

YEAR Farne Isle Fast SW Wales Donna Helms- Loch E. nan Shet- S. Ron- Islands of Castle Eng- Nook dale Eriboll Ron, land aldsay May land Tongue (Orkney) 1956 751 ...... 1957 854 ...... 1958 869 ...... 1959 898 ...... 1960 1020 ...... 123 1961 1141 ...... 152 1962 1118 ...... 1963 1259 ...... 1964 1439 ...... 115 1965 1404 ...... 74 1966 1728 ...... 107 1967 1779 ...... 132 1968 1800 ...... 152 1969 1919 ...... 127 1970 1987 . . . . 15 . . . . 103 1971 2041 . . . . 1 . . . . 148 1972 1617 . . . . 0 . . . . . 1973 1678 . . 107 . 0 . . . 578 123 1974 1668 ...... 136 1975 1617 ...... 197 1976 1426 ...... 160 1977 1243 . . . 645 . . . . 700 156 1978 1162 ...... 169 1979 1320 300 ...... 164 1980 1118 499 ...... 140 1981 992 505 . . . 34 . . . . 82 1982 991 603 . . . 43 . . . . 103

33 SCOS 02/2

Table 4 continued. Pup production estimates for other sites routinely monitored.

YEAR Farne Isle Fast SW Wales Donna Helms- Loch E. nan Shet- S. Ron- Islands of Castle Eng- Nook dale Eriboll Ron, land aldsay May land Tongue (Orkney) 1983 902 336 ...... 1984 778 517 . . . 30 94 406 . . . 1985 848 810 . . . 53 . . . . . 1986 908 891 . . . 35 . . . . . 1987 930 865 . . . 72 . . . . . 1988 812 608 . . . 54 . . . . . 1989 892 936 . . . 94 280 666 . . . 1990 1004 1122 . . . 152 . . . . . 1991 927 1225 . . . 223 321 . . . 241 1992 985 1251 . 1308 200 225 612 . . 246 1993 1051 1454 . 1372 205 . 700 . . 244 1994 1025 1325 . 1350 302 . 700 . . 258 1995 1070 1353 . . 334 300 516 . . 250 1996 1061 1567 . . 310 300 726 . . 250 1997 1284 1796 236 . . 382 523*” 719 . . 250 1998 1309 1968 273 . . 439 . 649 200 . 250 1999 843 1766 268 . . 503 . (422)” (83)” . . 2000 1171 2133 381 . . 618 . 670 235 . . 2001 1247 1932 321 . . 634 676*” . . . . * Includes pups on Berridale beaches. ” One flight only

34 SCOS 02/2

SCOS 02/2 ANNEX III The Status of British Common Seal Populations Callan Duck & Dave Thompson

1. Common seals surveys in eastern England 2001 In 1988, the numbers of common seals in The Wash declined by approximately 50% as a result of the phocine distemper virus (PDV) epidemic. Prior to this, numbers had been increasing. Following the epidemic, from 1989, the area has been surveyed once or twice annually in the first half of August each year (Figure 1, Table 1). One complete aerial survey of common seals was carried out in Lincolnshire and Norfolk during August 2001 (Table 1). The count for The Wash (3194) was the highest ever recorded. It was 15% greater than the mean of the 2000 counts (2,778) and 5% greater than the higher of the 2000 counts (3029). The average annual rate of increase in the number of seals counted in The Wash since 1989 is 6.3% (SE = 0.60%). This is significantly greater than the average annual rate of increase between 1968 and 1988 of 3.5% (SE = 0.29%). The 2001 count of common seals in The Wash exceeded the 1988 pre-epidemic count by 5%. It has taken 13 years for the population to recover from the effects of the PDV epidemic. This is in contrast to populations on the east and south sides of the North Sea, which recovered more rapidly and were similar to or exceeded their pre-epidemic levels by 1996. The 2001 counts at Blakeney Point and Donna Nook were lower than those in 2000 (Table 1), by 14% and 40% respectively. However, the average annual rates of increase in the number of seals counted at these sites since 1989 (16% (SE=3%) and 20% (SE=4%) respectively) are higher than in The Wash. Overall, the English East coast population has been increasing at an average annual rate of 7.4% (SE=0.5%).

2. Common seals in Scotland In August 2001, the coasts of Orkney and Shetland were surveyed for common seals. These are part of the third round Scotland survey and were part funded by Scottish Natural Heritage. The 2001 counts for both areas were lower than the two most recent previous counts in 1993 and 1997. In Shetland the 2001 count was 18.5% lower than the 1997 count and in Orkney the count was 9% lower. However, counts in both these regions lay within the range of the counts made since 1989. Other data for the University of Aberdeen has shown declines in common seal abundance in some areas of Orkney.

Location 1989 1991 1993 1997 2001 Orkney 7137 7873 8523 7752 Shetland 4794 6224 5991 4883

3. Minimum estimate of the British common seal population The most recent minimum estimate of the number of common seals in Scotland is 30,196 from surveys carried out in 1996, 1997, 2000 and 2001. The most recent minimum estimate for England is 4,409. This comprises 4,274 seals in Lincolnshire and Norfolk in 2001 plus 135 seals in Northumberland, Cleveland, Essex and Kent between 1994 and 1997 and an estimated 20 seals from the south and west coasts.

35 SCOS 02/2

Counts by region are given in the Table below for the periods 1996-1997 and 1996-2001. These are presented as the most recent counts for each region during the specified period. Where multiple counts were obtained in the most recent surveys, the mean values are presented.

Region 1996-2001 Shetland 4,883 Orkney 7,752 Outer Hebrides 2,413 Highland (Nairn to Appin) 6,291 Strathclyde (Appin, Loch Linnhe to Loch Ryan) 7,909 Dumfries & Galloway (Loch Ryan to English Border 6 at Carlisle) Grampian (Montrose to Nairn) 126 Tayside (Newburgh to Montrose) 165 Fife (Kincardine Bridge to Newburgh) 611 Lothian (Torness Power Station to Kincardine Bridge) 40 Borders (Berwick upon Tweed to Torness Power 0 Station) TOTAL SCOTLAND 30,196

Blakney Point 772 The Wash 3,194 Donna Nook 233 Scroby Sands 75 Other east coast sites 135 South and west England (estimated) 20 TOTAL ENGLAND 4,429 TOTAL BRITAIN 34,625

4. Common seal surveys proposed for 2002 Common seals in the Outer Hebrides, Northern Island and on the east coast between the Moray Firth and Thames estuaries will be surveyed in August 2002. The outbreak of a new PDV epidemic in the southern Baltic and Waddensee populations in 2002 makes it imperative that the whole of the East coast of the UK is surveyed in advance of any large scale mortality. These surveys will be part funded by SNH, Northern Ireland Office and DEFRA and these surveys have been completed

36 SCOS 02/2

Table 1. Numbers of commons seals counted on the east coast of England since 1988. Data are from fixed-wing aerial surveys carried out during the August moult.

Date of survey 13.8.8 8.8.89 11.8.9 2.8.91 1.8.92 8.8.93 6.8.94 5.8.95 2.8.96 2.8.97 7.8.98 3.8.99 4.8. 00 4.8.01 8 0 12.8.8 11.8.9 16.8.9 12.8.9 15.8.9 8.8.97 14.8.9 13.8.9 12.8.0 9 1 2 4 5 8 9 0 Blakeney Point 701 - 73 - - 267 - 438 372 250 535 715 895 772 307 - 217 196 392 371 738 602 dist. The Wash 3087 1531 1532 1226 1724 1759 2277 2266 2151 2561 *2367 2320 2528 3194 1580 1551 1618 1745 1902 2360 2381 2474 3029 Donna Nook 173 - 57 - 18 88 60 115 162 240 294 321 435 233 126 - - 146 36 262 201 286 345 Scroby Sands ------61 - 51 58 52 69 84 75 - - - - 49 72 - 74 9 The Tees ------35 - - - - - Holy Island ------(Northumberland) - - - 13 - 12 - - 10 Essex & Kent ------90 ------

* One area used by common seals was missed on this flight (100 – 150 seals); this data point has been excluded from analyses

37 SCOS 02/2

Counts of common seals in The Wash in August

4000

3000 epidemic

2000 Number of seals 1000

0 1965 1970 1975 1980 1985 1990 1995 2000 Year

Figure 1. Counts of common seals in The Wash in August. These data are an index of the population size through time.

38 SCOS 02/2 SCOS 02/2

ANNEX IV

Population dynamics of grey seals

I.L. Boyd

Sea Mammal Research Unit, Gatty Marine Laboratory, University of St Andrews

Introduction The management of the UK grey seal population requires the development of robust models of the underlying processes driving the dynamics of the population. Grey seal population dynamics do not differ fundamentally from those of other wildlife populations in terms of the underlying processes of birth, death, immigration and emigration. The uniqueness of the problem associated with grey seal populations comes in terms of the detailed features of the uncertainty surrounding the vital rates and the data that is available to allow an assessment of the quality of the model. Apart from the need to develop a population dynamics model as a management tool for grey seals in the UK, there is a requirement to derive an estimate of total population size for grey seals from the annual measures of pup production obtained from aerial surveys of the major breeding colonies. At its simplest level, this could be done by applying a simple multiplier to the pup production based upon our knowledge of the equilibrium age structure. For some purposes this could be a defensible approach but where there is the possibility of defining age structures with greater accuracy then it would appear to be reasonable to use these to increase the accuracy of estimates of total population size. The recent operational model of the UK grey seal population (Hiby & Duck in press) had specific characteristics that made it difficult to implement in a contemporary context. It was not designed to cope with a population that was showing density-dependent changes and it was not formulated to examine changes at the level of sub-sections of the UK population. Recent data from estimates of pup production (SCOS 2002) suggest that density-dependent processes may be operating in some parts of the population. In addition, staff changes at SMRU mean that this model can no longer be supported as an interactive tool that can be used to address the requests for advice from the Scottish Executive. These requests have placed greater emphasis on future trends in the populations which the model was not formulated to provide. Two new approaches have been taken at SMRU. These are conceptually similar but differ in their statistical rigour. The more rigourous approach, which is spatially explicit and will eventually operate at the level of individual colonies, includes density-dependence as a factor affecting juvenile survival. This approach is described in the accompanying information paper (Newman et al. SCOS 2002). The intention is to develop the model described by Newman et al. (SCOS 2002, Annex V) as the main population dynamics model for grey seals. However, this requires further work and, for the purpose of the present advice from SCOS we will retain the output from the Hiby & Duck model. The second approach, which is outlined here mainly as an illustration, considers each of the four main regions of the UK grey seal population as separate entities and attempts to simulate the pup production trajectory in each of these from simple prior distributions of vital demographic rates. However, both approaches suffer from the same inherent uncertainties in the empirical data and from the fact that pup production alone is insufficient to provide robust estimates of all demographic variables. An objective of this paper is to describe these uncertainties and their possible consequences.

Data sources Estimates of pup production in the UK grey seal population are mainly obtained from annual aerial surveys of all the main breeding sites. The methods and results are described in SCOS 2002 ANNEX II. Many of the estimates of vital rates in UK grey seals derive from studies carried out in the 1970s and 1980s.

39 SCOS 02/2 Survival rates for adult grey seals have been estimated using data from adult female grey seals culled as breeding animals at the Farne Islands during 1972 and 1975 (Harwood & Prime 1978). The age structure of these animals suggested a mean annual survival rate of 0.94. Subsequent samples of grey seals from both the Hebrides and, outside the breeding season in the North Sea (Boyd 1982) supported a relatively high survival rate because of the regular presence of seals that were 30+ years of age. Juvenile survival rates have been more difficult to estimate. Harwood and Prime (1978) derived these for grey seals at the Farne Islands by balancing the pup production with the adult age structure and productivity. They concluded that post-weaning first year survival was likely to be about 0.66. Subsequently, Hall et al. (2001) used mark-recapture analysis to provide an estimate of 0.62 (SEM=0.155) for the same parameter in females and 0.19 (SEM=0.084) in males of mean mass or condition at weaning. Fecundity rate could not be estimated from seals shot at breeding colonies during the 1970s because these were, by definition, already breeding. A subsequent sample of seals mainly from the Farne Islands taken outside the breeding season during the early 1980s provided an estimated pregnancy rate of 0.94 amongst sexually mature adult females and showed that sexual maturity occurred mainly at age 5 with a range from 3 to 7 years of age (Boyd 1985). However, a sample from the Hebrides obtained in 1978 gave a pregnancy rate of 0.83. More recent longitudinal mark-recapture studies at North Rona (Pomeroy et al. 1999) gave a natality rate of 0.80-0.97 amongst a group of females that were breeding regularly between 1979 and 1995 at North Rona (Hebrides). These estimates are similar to those used for examining grey seal population dynamics in the western Atlantic (Mohn & Bowen 1996).

Uncertainties associated with estimations of vital rates The uncertainty in the estimates of pregnancy rate based upon mark-recapture studies (Pomeroy et al. 1999) was similar to that for estimates obtained from cross-sectional sampling (Boyd 1985). The true fecundity rate probably lies between 0.80 and 0.95, although as suggested by Boyd(1985) it is possible that some sampling regimes over-estimate fecundity so there could be an upward bias associated with some current estimates of fecundity. The results of Hall et al.’s analysis (2001) suggest that the first year survival rate of females could be in the range 0.3-0.9. Overall, the empirical estimates of the vital rates of the UK grey seal populations suggest a degree of consistency between studies. In addition, they are consistent with the dynamics of a population that has been increasing at the rates observed during the past 30 years. However, there are few quantitative data about how density-dependence influences these vital rates and there are also few data, beyond anecdotal observations, to quantify the contribution of movement of grey seals between different regions. Hall et al. (2001) showed that first year survival was a function of the condition of pups at weaning which is, in turn, a function of maternal condition (Pomeroy et al. 1999). This suggests that first year survival should be responsive to the amount of food available to mothers in the previous year. Therefore, there is some evidence that both first year survival and fecundity will have a density-dependent component depending upon food availability. It has not been possible to determine the extent to which post-first year survival is density-dependent. In addition, we cannot say what the relative sensitivity to density will be amongst these vital rates or what measure of density is likely to be most important. Newman et al. (SCOS 2002) have investigate the effects of density dependent first year survival and recruitment to the local breeding population. Despite the internal consistency in the empirical measures of the vital rates, there is insufficient precision in these to allow investigation of the processes underlying the spatial and temporal variations currently observed in pup production or to investigate how these relate to total population size.

Density dependence: uncertainties and their consequences There is almost no information about how the vital rates respond to density in the UK grey seal population. Dispersal appears to occur as a function of density, at least on a local scale (Boyd 2002, Gaggiotti et al. 2002), and as a result of disturbance (see later) but it is not clear how much dispersal occurs between the regions of the population and how much this is driven by local density. For the purpose of this analysis,

40 SCOS 02/2 dispersal is assumed to be insignificant or, if it does occur, it is assumed that it will be contained within the variability in survival rate. The way in which density-dependence is factored into the assessment of grey seal population dynamics will influence the interpretation of the total population size derived from the estimates of pup production. This sensitivity is illustrated in Figure 1a. This shows that if changes in pup production are the result of adjustments in the fecundity rate then the estimated total population is considerably larger than if it is the result of changes in survival rates. The illustration in Figure 1a is for the Outer Hebrides and this shows that the total female population could be 30-40% larger if fecundity rather than survival are responsible for changes in the observed pup production. An additional feature of the population if it is controlled only by changes in fecundity is that it takes a very long time to stabilise after pup production has stabilised. This is illustrated in Figure 1b where the pup production for the Outer Hebrides population has been held at 2001 levels until 2015. Even after this time, the adult female population continues to increase, albeit at a slower rate than before.

a 40000

35000 Fecundity 1st year surv. 30000

25000 1+ surv.

20000

15000

Number of females 10000

5000 1965 1970 1975 1980 1985 1990 1995 2000

Projection 50000 b

40000

30000 Females

20000 Pups 10000

Number of females or pups Number 0 1960 1970 1980 1990 2000 2010 2020 Year

Figure 1. Investigation of the effects of uncertainty in which demographic variable is responsible for regulating the population. (a) Illustration of the effects of incorporating density dependence in different vital rates on total population size. The figure illustrates the data from the Outer Hebrides. (b) Pup production and total female population for the Outer Hebrides projected forward through time. The pup production has been held constant at 2001 levels after 2001.

Stochastic simulation of the population Consequently, the largest problems faced when modelling the UK grey seal population are the uncertainties within the vital rates and, consequently, also in the starting age structure. The approach

41 SCOS 02/2 described here attempts to use the time-series of pup productions as a way of exploring the distribution of the vital demographic rates and, therefore, the age structure. One approach to simulating the UK grey seal population is to develop a stochastic model that, as far as possible, relaxes assumptions about the precision with which the vital rates and starting age structure are known but which makes maximum use of the estimates of pup production which are the most precisely known population variables. This approach considers the UK grey seal population as 4 independent spatial units, Inner Hebrides, Outer Hebrides, Orkney and North Sea and movement between these is assumed to be insignificant. A fully spatially explicit model is described briefly in the accompanying information paper (Harwood & Newman SCOS 2002). For the purpose of the present simulation, however, the aim is to find the combination of fecundity rate, first year survival rate and post-first year survival rate that explain the variation in pup production. This simulation makes no explicit attempt to model the density-dependent process because this should be an emergent property of the simulation. Although the three vital rates being modelled are partly confounded, their effects can be partially disentangled by fitting each to several years’ pup production . Each demographic variable has a different effect upon the trajectory of the pup production, especially given the underlying history of the population expressed in the form of the age structure. Therefore, in order to fit to the pup production, there will be a most likely combination of fecundity, first year survival and post-first year survival that explains the trajectory of the pup production. Critical tests for this approach are whether, overall, it settles upon estimates of the three vital rates being considered that are consistent with the empirical observation and whether it is able to resolve specific events in the management history of grey seals at the Farne Islands. In particular, this includes culls targeted on breeding adult females during 1972 and 1975.

Implementation of the simulation The study was constructed as a Leslie matrix simulation. The fecundity rate, first year survival rate and post-first year survival rates were given prior distributions that were uniform. The only constraints on these distributions were that fecundity rates (female pups per adult female) were <0.5 and survival rates were <0.98. Age at sexual maturity was fixed at 5 years with a standard deviation of 0.25 years. The initial age structure was estimated using vital rates drawn from these uniform distributions and the modelled pup production for the population was allowed to converge with the starting pup production measured during the first survey in the time-series for each population. Multiple projections (10,000) of the population were then carried out for each starting age structure based upon the prior distributions of the vital rates. The simulation process involved stepping forward through the pup production estimates seeking the maximum likelihood values for the three vital rates at each step. Maximum likelihood was estimated based upon the fit to the pup productions for the following 5 years with the highest weights given to the current and most recent years. This helped the simulation to distinguish changes in pup production caused by changes in post-first year survival from changes in first year survival or fecundity.

Testing the simulation A critical test of the simulation as a method of investigating the changes occurring in vital rates is whether it is able to detect the occurrence of the culls that took place in the Farne Islands population during the 1970s and to represent these as changes in the adult female survival rate. The main culls that occurred took place in 1972 when 554 breeding females were killed and then in 1975 when 482 breeding females were killed. A further 92 females, most of which were adults, were killed outside the breeding season between during 1979 and 1981. As can be seen from Figure 2, all these events are detectable in some form in the post-first year survival rate (Figure 2). There was an overall decline in post-first year survival between 1971 and 1978 with particularly low levels of survival in 1971 and 1975. It is not clear why the cull that took place in 1972 is represented in the estimate for 1971. However both events show declines in the survival rate of about 0.1. There were about 4000 females in the population at the time (Fig. 2) which gave an added

42 SCOS 02/2 component of mortality to females of 0.12. The decline in the survival rate derived by the simulation is, therefore, close to the expected level. In addition, management at the Farne Island through the 1980s involved disturbing females that attempted to pup on specific islands. This may account for the reduce fecundity rate during the 1980s. Disturbance due to the culls in the 1970s may also have led to the reduced apparent survival rate during this time. In fact, some of this may have occurred as emigration. For example, the grey seal colony on the Isle of May was established at this time and some individuals may have moved even further. The simulation was free to select vital rates from a very wide range of possibilities. However, the posterior distributions (Fig. 2) lay within the ranges expected from empirical observations. As illustrated for the North Sea (Fig. 2), and as might be expected, there was greatest uncertainty in the first year survival rate.

a 16000 Projection

14000

12000

10000 Total females 8000

6000

4000

Number of pups or adults or pups Number of 2000 Pups

0 1950 1960 1970 1980 1990 2000

b 1.0 Post 1st year survival 0.8

1st year 0.6 survival Rate

0.4 Fecundity

0.2 1950 1960 1970 1980 1990 2000 Year

Figure 2. Changes in the number of pups and post-first year females in the North Sea section of the UK grey seal population (a). Panel b shows the variability and posterior distributions of vital rates through time for the same section of the population. Arrows show when the major culls took place at the Farne Islands in 1972 and 1975. See legend to figure 2 for a description of the projection.

Projection of the population Since we know very little about how density dependence will operate there are no grounds for forward projection of the population other than to assume that the recent population history is indicative of future trends. Even if we were to assume that first year survival or fecundity were likely to be affected by population density to a greater extent than post-first year survival, we have no information about the carrying capacity for each of these vital rates.

43 SCOS 02/2 In this case, population projection has been carried out by running the simulation forward for 5 years using bootstrapped values of the posterior distributions of vital rates for the previous 5 years. Confidence limits were calculated by repeating the process for 1000 sets of simulations. The results are shown in Figure 3. Projection suggests that the populations in the Hebrides are most likely to stabilise over then next 5 years whereas those in Orkney and the North Sea will continue to increase. In addition, overall, the UK grey seal population is likely to continue to increase for the foreseeable future. However, the confidence intervals on these projections are very large. For example, by 2006, the number of female grey seals associated with regularly surveyed sites could lie between 63,000 and 123,000 individuals (Fig. 3). This compares with the current range of 65,000-86,000 for the same section of the population.

120000 a Projection

100000 Total females

80000

60000

40000

20000 1950 1960 1970 1980 1990 2000 2010

Projection 70000 b

60000 Number ofNumber females 50000 Orkney

40000

30000 O.Hebrides 20000 North Sea 10000 I.Hebrides 0 1950 1960 1970 1980 1990 2000 2010

Year Figure 3. The total number of females (a) in the UK grey seal population and (b) in each region. The mean estimate is shown for each year in which there is an estimate of pup production and this is shown with the standard deviation. In addition, the result of population projections to 2006 are shown using the distribution of vital rates from each section of the population over the period 1997-2001. The projection for the total population is the sum of the projections for each region.

44 SCOS 02/2 Estimating population size associated with sites surveyed infrequently Pup counts are occasionally available from locations that are not part of the annual aerial surveys. In the order of 5,000 pups are born at these sites (SCOS 2002 ANNEX II). In order to estimate the total female population associated with these sites, a multiplier can be applied to the number of pups based upon the age structures calculated in the simulations of the population in the 4 main regions. The mean multipliers for the 4 regions were: North Sea, 2.05 (SD=0.24); Orkney, 2.05 (SD=0.19); Outer Hebrides, 2.27 (SD=0.11); Inner Hebrides, 2.01 (SD=0.23). Since the number of seals at many of the sites not surveyed frequently is not changing rapidly (e.g. in SW Britain), the multiplier for the Hebrides is likely to be the most appropriate. This suggests that there are 10,000-12,500 female seals associated with these additional sites.

Total population size According to this analysis, the total female population size at regularly surveyed sites in 2001 was between 65,000 and 94,000 individuals. The historical total female population size and the numbers of female seals associated with each region are shown in Figure 3. Including sites surveyed less frequently puts the total female 1+ population at 75,000-106,500. We cannot model the population of males in the same way as for females but on the assumption that the male population is about 60% of the female population the total population of grey seals associated with regularly surveyed sites is between 122,000 and 168,000 with a median of 144,000. The number of seals in the grey seal population together with the annual percentage change in the pup production and the total population are shown in Figure 4.

180000 15 change percentage Annual % change (pups) % change (total) 160000 10 140000 5 120000

100000 0

Number of seals No. of seals 80000 -5 60000 1985 1990 1995 2000

Year Figure 4. The estimated total number (±SD) of seals in the UK grey seal population and the annual percentage change in the population (solid line) and the pup production (dotted line).

Comparing the results of different methods of estimating total grey seal population size

The results from the methods of estimating total population size used by Hiby and Duck (in press) are shown in Table 1.

45 SCOS 02/2 Table 1. Estimated size of the grey seal population associated with all major, annually monitored, breeding colonies in Scotland and eastern England, except Loch Eriboll, Helmsdale and Shetland. Estimates refer to the number of seals aged 1 and over at the start of the 2001 breeding season.

Year Estimated pup Total female Total population production population (survey) Hiby & Duck Hiby & Duck 1984 14,992 25,824 44,489 1985 16,265 27,274 46,956 1986 17,796 28,983 49,923 1987 19,035 30,804 53,090 1988 18,071 32,674 56,332 1989 19,926 34,511 59,481 1990 21,093 36,363 62,622 1991 23,815 38,361 66,017 1992 27,075 40,570 69,795 1993 28,338 42,878 73,740 1994 29,018 45,349 77,982 1995 30,932 47,853 82,255 1996 33,426 50,561 86,895 1997 32,771 53,469 91,893 1998 35,680 56,497 97,082 1999 33,103 59,648 102,546 2000 36,915 62,905 108,027 2001 36,920 66,522 114,244

Table 2. Pup production and associated population size for the main, annually monitored grey seal breeding colonies in 2001.

Location 2001 pup production Change in pup Total 2001 population production from 2000 (to nearest 100) Inner Hebrides 2,938 -9% 9,100 Outer Hebrides 12,325 -8% 38,100 Orkney 17,523 +10% 54,200 Isle of May & Fast 2,253 -10% 7,000 Castle Farne Islands 1,247 +6% 3,900 Donna Nook 634 +3% 2,000

Subtotal 36,920 0% 114,200

SW England & Wales 1,500 4,600 All other colonies 3,500 10,800

Total 41,920 129,700

The estimate of the total grey seal population at annually monitored colonies was 114,000 in 2001. The 95% confidence limits for the entire female population are within 14% of the point estimate (57,000- 75,000). The table above has been generated using the same method as in previous years. Figure 5 shows the population trajectories together with the predictions for the years 2002-2007. Table 2 shows the total population sizes associated with each section of the grey seal population based upon the Hiby & Duck model. The results of the analyses from the new method described by Newman et al. together with those from the present analysis are described in Table 3. This shows that the point estimate for the population

46 SCOS 02/2 size was 107,000 using the Newman method and 127,000 using the current analysis. There is however, considerable overlap in the upper and lower and lower bounds. Overall, therefore, these three methods give estimates of total population size in the annually monitored portion of the population that are not significantly different from each other. Nevertheless, the greater level of statistical rigour in the models of Hiby & Duck and Newman et al. means that greater weight should probably be placed upon their estimates than that of the current illustration.

Grey seal population trajectories at annually monitored colonies (Hiby method)

180000 Total population Total female population 160000 Estimated pup production (model) Estimated pup production (survey) 140000 Predicted total population Predicted female population 120000

100000

80000 Number of seals 60000

40000

20000

0 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 Year

Figure 1. Grey seal population trajectories at the annually monitored colonies around Britain. These data have been generated using the Hiby & Duck method and can be compared with results presented to SCOS in previous years. Predictions for the years 2002 – 2007 are included.

Table 3. Comparison between estimates for 2001 from Newman et al. (SCOS 2002) and the simulation results for 2001 described here for the adult male plus female population size at annually surveyed sites. Newman et al. Current analysis Estimate Lower Upper Estimate Lower Upper North Sea 11,741 7,586 18,080 14,214 9,771 18,658 Orkney 58,498 31,767 84,819 58,360 44,113 72,604 O Hebrides 29,494 20,404 54,640 43,910 26,856 60,964 I Hebrides 7,138 5,320 12,422 10,582 7,273 13,890 Total 106,871 65,077 169,961 127,066 88,013 166,116

Future model development

Unlike the Hiby & Duck approach the Newman model includes density-dependence in at least one demographic variable and is being developed as a tool for modelling the dynamics of grey seal populations in the UK. Because of the way in which it has been written and also because of changes in personnel at

47 SCOS 02/2 SMRU, the Hiby & Duck model can no longer be supported. The intention is, therefore, to move to an alternative model, such as the Newman model, as soon as possible.

References

Boyd, I.L. (1982). Reproduction in grey seals with reference to factors affecting fertility. PhD Thesis, Cambridge University. Boyd, I.L. (1985). Pregnancy and ovulation rates in grey seals (Halichoerus grypus) on the British coast. Journal of Zoology, London 205: 265-272. Boyd, I.L. (2002). The measurement of dispersal by seabirds and seals: implications for understanding their ecology. In Dispersal Ecology, (eds) J.M. Bullock, R.E. Kenward & R.S. Hails, pp72-88 . Blackwell, Oxford. Gaggiotti, O.E., Jones, F., Amos, W., Harwood, J. and Nichols, R.A. (2002). Patterns of colonization in a metapopulation of grey seals. Nature 416: 424-427. Hall, A.J., McConnell, B.J. & Birker, R.J. (2001). Factors affecting first-year survival in grey seals and their implications for life history strategy. Journal of Animal Ecology 70: 138-149. Harwood, J. & Prime, J.H. (1978). Some factors affecting the size of British grey seal populations. Journal of Applied Ecology 15: 401-411. Hiby, L.R., & Duck, C.D. (in press) Point and interval estimates of the size of the British grey seal Halichoerus grypus population and their implications for management. J. appl. Ecol. Mohn, R. & Bowen, W.D. (1996). Grey seal predation on the eastern Scotian Shelf: modelling the impact on Atlantic cod. Canadian Journal of Fisheries and Aquatic Sciences 53: 2722- 2738. Pomeroy, P.P., Fedak, M.A., Rothery, P. & Anderson, S. (1999). Consequences of maternal size for reproductive expenditure and pupping success of grey seals at North Rona, Scotland. Journal of Animal Ecology 68: 235-253.

48 SCOS 02/2 SCOS 02/2 ANNEX V Information papers provided for SCOS

1. Newman, K., Thomas, L. & Harwood, J. (2002) Estimates of the current size of the different components of the British grey seal population based on a spatially-explicit state-space model.

2. Hall, A.J. (2002) Phocine Distemper Epidemic in European Harbour Seals, 2002. SCOS information paper.

3. Matthioloulos, J. Estimating spatial usage by marine mammals. SCOS information paper.

49 SCOS 02/2

SCOS Inf. Paper 02/01

Estimates of the current size of the different components of the British grey seal population based on a spatially-explicit state-space model.

Ken Newman, Len Thomas and John Harwood

Introduction and Methods We have developed a spatially-explicit model of the dynamics of the British grey seal population using an application of the state-space framework (Akoi 1990) developed by Buckland et al. (submitted) for population dynamic applications. One characteristic of the state-space approach is that the true, but unknown, state of the population is modelled by a state process which is linked to survey data by a separate observation model. The state process model incorporates stochasticity in demographic rates. Posterior distributions for parameter values and population sizes can be generated using numerical Bayesian approaches. In this case, we have used a Markov chain Monte Carlo algorithm. Our model departs from a strict state-space approach in that the state process is 6th order, rather than 1st order, Markovian. This reflects the fact that animals are not recruited to the breeding population until their 6th birthday.

Our state process model includes functions for density-dependent juvenile survival and density-dependent migration. Although we believe that these processes probably operate at the level of individual colonies, we have not yet been able to implement a model at this spatial scale. Instead, we have aggregated colonies into “super-colonies” that are equivalent to the regional divisions traditionally used to summarise the results of SMRU’s surveys: North Sea (Isle of May + Fast Castle + Farnes + Donna Nook), Inner Hebrides, Outer Hebrides, and Orkney. Juvenile survival and migration in a particular year are assumed to be related to the total number of pups in each super-colony. Adult survival (applied to all 1+ animals) and fecundity are assumed to be the same for all super-colonies. The number of adults surviving each year was assumed to be a binomial variable.

Juvenile survival (фjct) at super-colony c in year t is described by a Beverton-Holt function of the form:

where α is the maximum value for juvenile survival, and 1/βc reflects the carrying capacity of all the colonies within super-colony c, following Harwood (1981) The number of pups surviving to the end of the first year was also assumed to be a binomial variable whose probability was determined by the above expression multiplied by adult fecundity and the number of 6+ animals (i.e. those six years and older).

We have assumed that only females recruiting to the breeding population for the first time will migrate, and that the probability of their moving from their natal super-colony to another super-colony is a function of the difference between juvenile survival at the two sites, and distance between super-colonies. Colony- specific multinomial distributions are assumed for movement from one super-colony to any other (including to itself, i.e., not moving), and the proportions are based upon the extension of the logit transform for a binomial proportion to multinomial proportions:

50 SCOS 02/2

dcb is the average distance between all pairs of colonies within super-colonies c and b, and Ib=c is an indicator for site b equalling site c.

In effect, this imposes a form of density-dependent fecundity, within a super-colony, on animals breeding for the first time. The only published evidence of density- dependent fecundity variation in phocid seals (Bowen et al 1981) involved animals in this age category.

We fitted this model to the “most consistent” time series of pup production estimates from 1984-2001, although the first six data points were used to establish the initial age structure of the population. We initially assumed that pup production estimates had a coefficient of variation (CV) of 5%, in line with the figure quoted by Duck et al. (submitted). However, this resulted in some very erratic population trajectories, particularly for the later years in the time series. We therefore repeated the fitting procedure using a CV of 10%.

Results Figure 1 shows the median trajectory for 50,000 permutations drawn at random from the prior distributions and then filtered according to their contribution to the overall likelyhood. Figure 2 shows the posterior distributions for the demographic parameters.

Overall, the model provides a reasonable representation of the observed changes in pup production over the last decade. However, it cannot completely capture the very rapid levelling off of pup production in the Inner and Outer Hebrides in the mid-1990s, or the rapid increase in pup production observed in Orkney at about the same time. It may be possible to reconcile these inconsistencies if older females are also allowed to migrate.

Estimates of the total population of 1+ animals at the start of each breeding season in all the super-colonies combined are approximately 10% higher than those obtained using the method of Hiby et al (in press), which assumes that the population is increasing exponentially. Table 1 shows the median, and upper and lower 95th percentiles for the posterior distributions of population size in 2001. Following Hiby et al. (in press), we have assumed that there are equal numbers of male and female seals up to the age of 5 years and that beyond this age males suffer higher mortality, so that male numbers are 60% of females. The difference between our figures and those obtained using the method of Hiby et al. is primarily due to the lower estimates of fecundity (0.921 vs 0.954) and adult survival (0.94 vs 0.97) we have obtained. Juvenile survival in the Inner and Outer Hebrides must have fallen to a relatively low value (0.27 and 0.24 respectively) if it is to account for the observed slowing down of pup production. We estimate that juvenile survival in Orkney is 0.57, and 0.43 in the North Sea. This compares with Hiby et al.’s estimate of 0.39 for the entire British population. Hall et al. (2000) estimated that the survival of female grey seal pups in the North Sea from weaning to age one was 0.62. Survival from birth to weaning varies from colony to colony, but estimated values are centred around 0.8, suggesting an overall first year survival of 0.5.

We can use the posterior distributions for the demographic parameters and population sizes to obtain projections of the population’s potential dynamics over the next five years. Figure 2 shows population trajectories until 2006 obtained by resampling from these distributions. We predict that the total size of populations associated with colonies in the Inner and Outer Hebrides will increase by less than 10% over

51 SCOS 02/2 this period, but population numbers in the North Sea and, particularly, Orkney will continue to increase by 5-7% annually. However, these are preliminary figures and there is substantial uncertainty associated with these projections.

Discussion Clearly, further work is required before this model can provide an entirely convincing description of observed changes in pup production in the different super-colonies. We will try to address the following issues over the next year:

• improvement of the model of migration to take account of additional data from mark-recapture, genetic and telemetry studies (this will be carried out by a PhD student funded under the NERC/EPSRC Environmental Mathematics and Statistics programme); • obtaining more realistic values for the uncertainty associated with estimates of pup production; • developing a methodology for obtaining colony-specific estimates of pre-weaning pup survival from aerial photographs; • investigating of alternative forms of density-dependence in survival and fecundity; • incorporating environmental stochasticity in adult survival and/or fecundity; • developing an equivalent model at the scale of individual colonies.

In addition, we will consider what needs to be done to obtain reliable estimates of the male component of the population.

52 SCOS 02/2 Table 1. Projections of the size of the various components of the British grey seal population at the start of the 2002 and 2006 breeding seasons. The 2.5th and 97.5th percentiles can be treated as the equivalent of upper and lower 95% confidence limits.

2.5th 97.5th 2.5th 97.5th percentile Median percentile percentile Median percentile “Super- 2002 2006 colony” North Sea 11,844 16,520 21,636 11,521 20,498 31,620 Inner 7,578 10,243 14,399 6,810 11,148 18,298 Hebrides Outer 27,410 37,719 52,615 23,216 40,724 65,694 Hebrides Orkney 53,286 75,102 95,556 56,696 99,234 151,637 100,118 139,584 184,206 98,243 171,604 267,249 Total

Figure 1. Predicted and observed trajectories for grey seal pup production at four British “super-colonies”. A. North Sea, B. Inner Hebrides, C. Outer Hebrides, D. Orkney

A. North Sea B. Inner Hebrides

C. Outer Hebrides D. Orkney

53 SCOS 02/2 Figure 2. Prior and posterior distributions for demographic parameters.

Figure 3. Population projections for four British grey seal “super-colonies”. A. North Sea, B. Inner Hebrides, C. Outer Hebrides, D. Orkney.

A. North Sea B. Inner Hebrides C. Outer Hebrides D. Orkney

References Akoi, M. 1990. State Space Modelling of Time Series. Springer-Verlag, Berlin. Bowen, W. D., Capstick, C. K. & Sergeant, D. E. (1981). Temporal changes in the reproductive potential of female harp seals (Pagophilus groenlandicus). Can. J. Fish. Aquat. Sci. 38, 495-503. Buckland, S.T., Newman, K.B., Thomas, L. & Kösters, N. (submitted) State-space models for the dynamics of wild animal populations. Statistical Science. Harwood, J. (1981) Managing grey seal populations for optimum stability. Pp 159-172 in: C.W. Fowler and T.D. Smith. (Eds) Dynamics of Large Mammal Populations. John Wiley & Sons, New York. Hiby, L.R., & Duck, C.D. (in press) Point and interval estimates of the size of the British grey seal Halichoerus grypus population and their implications for management. J. appl. Ecol.

54 SCOS 02/2 SCOS Inf. Paper 02/02

Phocine Distemper Epidemic in European Harbour Seals, 2002

Ailsa J. Hall Sea Mammal Research Unit, Gatty Marine Laboratory, University of St Andrews, St Andrews KY16 8LB

In May 2002 an unusually high mortality among harbour seals, Phoca vitulina breeding on the Danish island of Anholt in the Kattegat was reported. Some carcasses were recovered and post-mortem examinations confirmed the cause of death to be phocine distemper virus (PDV). Virological investigations determined there was a 97% identity between this virus and the 1988 PDV strain1. Since then the disease has spread from seals in the Danish and Swedish Kattegat and Skagerrak into the population inhabiting the Dutch and German Wadden Sea. Almost 3,500 dead seals have now been counted with the number still rising (Fig 1.). To date one case has been confirmed in a seal found on the Northern French coast near the Belgian border and three possible cases in the UK are being investigated at present.

A dedicated telephone line for reporting dead and sick seals has been set up at the Institute of Zoology in London, who will be coordinating the UK investigation. Scottish Agricultural College VIC in Inverness will carry out all post mortems on Scottish seals. SMRU will assist with disseminating information via an e-mail list to all interested governmental, NGO and scientific parties, will keep a dedicated web page updated weekly and ensure necessary data are collected for subsequent epidemiological modelling. Some post mortem, serology and virology may be undertaken if necessary.

1Jensen, T., van de Bildt, M., Dietz, H.H., Andersøn T.H., Hammer, A.S., Kuiken, T. and Osterhaus, A. (2002). Another phocine distemper outbreak in Europe. Science, 297, 209.

Fig. 1

European Seal Deaths May-August 2002

4000

3500

3000 Denmark

. Sweden 2500 Netherlands 2000 Norway France/Belgium 1500 Cumulative No Germany

1000 Total

500

0 14-Apr-02 4-M ay-02 24-M ay-02 13-Jun-02 3-Jul-02 23-Jul-02 12-Aug-02 1-Sep-02 Date

55 SCOS 02/2 SCOS Inf. Paper 02/03

Estimating spatial usage by marine mammals J. Matthiopoulos, SMRU

Over the last five years SMRU has been developing expertise, in modelling animal movement, specifically in relation to marine mammals.

As a result, a considerable amount of mathematical and numerical techniques have now been developed and organised into software libraries for ease of implementation and modification.

Specific applications currently include estimating the range of the population of British grey seals and North American steller sea lions.

Modelling the range of the British grey seal population.

Models of movement are closely interlinked with techniques of statistical estimation through the concept of model-supervised usage estimation developed within SMRU.

We use auxiliary information and mechanistic modelling to assist the estimation of spatial usage from sparse telemetry data.

These estimates are considerably more accurate than those obtained via traditional methods such as kernel smoothing.

Estimating spatial usage by the British grey seal population

56 SCOS 02/2 Discrepancies between the models’ predictions and the final estimate of spatial usage can be very informative.

In fact, these discrepancies can serve as the first step in a quantitative investigation of habitat preference by identifying which environmental attributes might be considered as covariates of usage.

Of course, a trained observer could identify such hotspots by looking at the raw data from a group of tagged animals. But how easy would it be to make quantitative statements about the entire population of animals just by eyeballing the data?

Hotspots of spatial usage in British grey seals. Light purple indicates that the animals used those parts of space less than originally expected. The gradient from green to red is used to show usage that exceeds prior expectations.

57 SCOS 02/2